The Digital Ouroboros: Social Media Is Eating Its Own Tail
For years, we have criticized social media as echo chambers that stifle critical thinking and fuel polarization and disinformation. But it turns out the problems they create are far greater than mere cognitive narrowing.
Today, the dominant conversation about AI on social media focuses primarily on quality. We complain about "AI slop": the endless stream of AI-generated, vacuous text and imagery polluting our feeds. While this is indeed annoying, what if it is also existentially dangerous?
The real crisis is not that our feeds are becoming uglier. The real crisis is that the infrastructure of social media, due to the total absence of human governance, has transformed from a marketplace for attention into a platform for economic and political extraction. We no longer just "use" social media; we are mined by it. We are the raw material; we are being extracted.
To understand the impact of this economic and political extraction, we must first quantify the pollution. A prediction that a major part of our online content would be synthetic by 2026 seemed alarmist just a few years ago. Today, this is no longer science fiction. Gartner predicts that by next year, at least a third of all online content will be synthetic. Internal analyses and market studies show that the tipping point is already behind us. On platforms like Meta, 41% of long-form posts were partially or fully synthetic as early as 2024. By mid-2024, over 15% of video uploads on TikTok already contained synthetic effects or AI voiceovers. We are now in 2026, and synthetic content is growing exponentially.
But why is this synthetic content so dangerous? Because it offers organized crime and shadow state actors a lever with no equal in human history. It is malicious leverage. Historically, structural disinformation or large-scale fraud required an organization and significant capital: expensive campaigns, call centers, and armies of humans working behind PCs. Today, the marginal cost to generate one fraudulent ad or one targeted piece of misinformation on social media is identical to the cost of generating ten million: virtually zero. A single actor can use AI to generate thousands of variants of scam or disinformation scenarios with a single click, each targeted at the specific psychological vulnerabilities of the platform's users. There is no longer a brake on production.
In this model, social media platforms are not neutral conduits. They are the highly efficient logistics partners of the underworld. They provide the targeting that allows bad actors to learn who is most susceptible to what, and they help optimize the message. Malicious actors utilize the platforms' own A/B testing tools to refine their narratives in real-time. The algorithm, blind to intent and purely focused on engagement, notices that an emotionally extreme fraudulent ad scores well and rewards it with more reach. This is an incentive trap: platforms do this because they want more content, which generates more clicks and time-on-site. This is exactly what these shadow actors want: massive volume to increase their surface area, maximizing clicks and eyeballs. Furthermore, there is the phenomenon of synthetic social proof. A malicious actor commands a swarm of bots to specifically like, share, and provide synthetic commentary on that content. The result is that the algorithm rewards this further, creating the illusion of virality.
The effectiveness is terrifying. Ordinary users can no longer distinguish synthetic content and artificial virality from the real thing. It is no longer a fair fight. The risk of being caught by the algorithm, the cost of being exposed as a fraudster, is nil. Economic and political extraction requires neither identity nor reputation; the actor only needs an API key.
This would be disturbing enough if the system were designed solely to steal money, if it were merely a scam economy. But the extraction is also political. The tsunami of "AI slop" flooding our information space and timelines is expertly designed to cause structural contradictions and systemic noise regarding how we perceive information, news, and facts. Through this flood, you eventually believe nothing. You mistrust everything. You lose faith in truth, in democracy, in institutions, in journalism. And that is precisely the ultimate goal.
The intentions of malicious actors and social media platforms are aligned: more attention leads to more money, which leads to more influence.Let’s follow the money. This is where the analysis becomes cold and transactional.In 2025, social media in Europe generated an estimated €45 billion in advertising revenue. It is estimated that roughly 10% of those revenue streams are directly linked to fraud, scams, and illegal trade. This means that an estimated €4.5 billion flows annually from the malicious shadow circuit directly into the books of the tech giants.The mechanism is as follows: shadow actors steal money from victims via crypto scams, romance frauds, fake web shops, and other forms of deception. A portion of that loot is immediately reinvested to buy new advertising space on Meta, X, TikTok, or Google. To the platform, this is legitimate revenue. The fraudster's Euro is worth just as much as the baker's Euro. A customer is a customer.
Here we see the perverse incentive in its naked form. Stricter controls, identifying the client or moderating content, cost the platform money, create friction, and thus lower revenue. Letting the synthetic stream run wild actually yields more money. The platform is, in effect, facilitating the laundering of criminal money through its own ad auctions. They allow shadow state actors to finance their hostile activities with money stolen from the population of their adversary. It is ingeniously diabolical.
The same infrastructure that robs your grandmother of her life savings is used by state actors to destabilize democracies. For a hostile state, the return on investment of social media is unbeatable. Why invest in expensive conventional weapons when, for a fraction of the price, you can demolish the societal cohesion of your opponent? By strategically deploying extreme sentiment, fueled by synthetic content and amplified by bot networks, polarization is driven to the breaking point. The algorithm acts as a catalyst here. It rewards anger, fear, and division because this generates engagement. Hostile states do not need to create the division; they only need to pay to let the algorithm do its work. Democracy is hollowed out from within, funded by advertising budgets. We are looking at a system that undermines its own right to exist.
In mythology, the Ouroboros is the snake eating its own tail, a symbol of the eternal cycle, but in this case, a symbol of self-destruction. The current model of social media is eating itself alive. Due to the dominance of synthetic content, user experience and trust are plummeting. The head of the algorithm is eating real people, real human attention, and real Euros out of the economy. To feed quarterly figures, the platform must continue to accept and monetize the cheap, synthetic mass. In the medium term, this is unsustainable, even for an untouchable tech giant.
The traditional Ouroboros symbolizes the positive cycle of life, death, and rebirth. A digital Ouroboros, however, may not follow a natural cycle. Driven by an insatiable craving for data and engagement, it may be doomed to ultimate self-destruction.
Let us ensure our knowledge of mythology serves us well. It is time for adversarial design in relation to social media. Everywhere around us, we see a shift in mindset; no one believes the fable of social media as a "connector of people" anymore. Let’s assume the system is being weaponized against us.Let’s demand transparency from platforms. Let’s follow the money. Who is funding the slop? Let’s make the platforms liable: if you create the snake, you have to pay the vet bill. No soft law; make the fines bigger than the profits. We must design rules to impose friction on social media. Make the abuse by the malicious actor more expensive than the gain they can possibly make from it. Impose verification for every economic entity that pays a platform. Impose IDs for every bot that interacts with a platform. Enforce a slowdown on the velocity of viral content until it is vetted.
We cannot stop social media from eating its own tail. But we can be the real human Ouroboros. Homo Social Media was born about 15 years ago; it is now facing death by a tsunami of bots. Another human will be born out of that demise, one who has learned to be skeptical of systems that can be captured and weaponized, simply because of our innate need to connect with one another.
Synthetic image/AI-generated
This blog is written by Patrick Lacroix in a personal capacity. AI tools are used for research, structuring, drafting and language support. All content is selected, verified, and edited by the author, who retains full editorial responsibility.